Create translation

post deprecated

https://api.macpaw.com/ai/v1/audio/translations

DEPRECATED. Translates audio into English using a unified OpenAI-compatible multipart interface.
Non-streaming only. Only OpenAI is supported — use openai/whisper-1.

Required fields (this endpoint):

Field	Description
`file`	Audio (or video) file to translate (`format: binary`).
`model`	Model id in `provider/model` format. Use `openai/whisper-1`.

OpenAI native requirements (Create translation):

Field	Required	Description
`file`	yes	Audio file (flac, mp3, mp4, mpeg, mpga, m4a, ogg, wav, webm).
`model`	yes	Must be `whisper-1` (on this endpoint: `openai/whisper-1`).

Optional parameters (OpenAI): prompt (English style/context hint), response_format
(json, text, verbose_json, srt, vtt), temperature.

The set of accepted optional fields and response fields may differ when additional providers are
supported. Provider-specific capabilities (including caching where supported) depend on the model —
use the model/info endpoint to check.

Recent Requests

Time	Status	User Agent
Retrieving recent requests…

Loading…

Body Params

Multipart form data. file and model are always required.

Optional fields follow the OpenAI Create translation
API.

Translation request. Required: file, model (openai/whisper-1).

See OpenAI Create translation
for full parameter details.

file

required

The audio or video file to translate into English.
Supported formats: flac, mp3, mp4, mpeg, mpga, m4a, ogg, wav, webm (see OpenAI docs).

model

string

required

Model id in provider/model format. Required. Only openai/whisper-1 is supported.

prompt

string

Optional English text to guide style, spelling, or domain vocabulary.

response_format

string

Output format: json, text, verbose_json, srt, or vtt.
Default json returns { "text": "..." }; verbose_json adds duration, language, segments.

temperature

number

Sampling temperature (0–1 typical). Higher values increase randomness.

Responses

200Success. Response body shape depends on response_format. Default json — { "text": "..." } with English translation. verbose_json — adds duration, language (always english), and segments[]. Response fields may vary when additional providers are supported.

400Bad request — invalid file or parameters

401Invalid token

402Insufficient credits

403Forbidden — authenticated but not allowed to access the requested model

429Rate limit exceeded

500Internal error